KG Balanced Dynahi build discussion thread

[jose]

Thanks Dr. Gilmore

 

I will check the cap. tomorrow. I suppose that it´s the correct value but...

 

Sincerely, I will prefer a simple error on the caps code.

 

We buy a big lot of Mospec and after "match", they are working fine on the Krell, but I will take note...

 

Thanks

[jose]

Dr. Gilmore  

 

You are right. For an error I mistook the cap. value for 0,2uF  

 

Now my Dynahj is flying    

 

 

are you sure those are 2pf caps, not 2uf caps. because that would definitely kill all the gain.

 

edit: yep 2uf would take the gain down to about -50db at 1khz.

 

 

 

 

 

 

 

 

I wish you a happy new year.

[kevin gilmore]

.2uf would be gain of -30db at 1khz.

but it would have had great low frequency response.

most of it subsonic.

 

happy new year.

[jose]

New dates:

 

1Khz, 1 Volt Rms to input

 

Out put: 17,138 Volt Rms

 

 

Thanks

[kevin gilmore]

unbalanced or balanced in, unbalanced or balanced out?

 

if you built it with standard parts, unbalanced in, unbalanced out, the gain should be 19.4db

Edited December 31, 2013 by kevin gilmore

[Torpedo]

AFAIK it's unbalanced in 1Vrms at 1KHz, the reading unbalanced out is 17,1 Vrms. Unsure if it's still 1KHz

 

Happy new year to all

[jose]

Hello

 

It´s balanced input and unbalanced output.

[jwzhan]

I just powered up my dynahi and used cheap apple buds for testing... The result is... weird...

 

There is a somewhat loud weird inconsistent crackling/humming noise for a brief 1 or 2 seconds immediatly following turn on when the earbud is plugged in.. and there is a sharp pitched noise about 1 second after turning off (not immediately, but occurs after 1 sec. as DC offset ramps up assumably.)

There is also a consistent light humming noise in the background which is likely due to how the amp is setup and wired at the moment.

 

Well... I dared to plug my DT880 250ohm (which is probably not a smart move) in there given the above result since the apple earbuds didn't blow up on me, and did play music when the amp was on. The initial crackling/humming noise is still there with similar volume. But the turn-off pitch is not there or lessened by a significant amount.

 

I have adjusted the amps so that DC offsets are under 100mV without the OPAMP in warm state. With OPAMP, the DC Offset is ±10mV. The voltages across the 2*20ohm resistors in all channels are 0.9V, which doesn't really matter....

 

I know that it takes time for the OPAMP (445) to start working, could it be the cause? It occurs in both amp boards. I'll have more detail tonight if I have more time to work on the amp.

Edited January 15, 2014 by jwzhan

[Pars]

If it were mine, I would get it so that DC offset is in the 1-2mV range WITHOUT the opamp in. Yours is much too high at 100mV. Are you using all 4 (or is it 8 for balanced?) pots? 620 or 680 ohm resistors in parallel with the pots?

[jwzhan]

I tried to get it as low as possible but it drifts quite a bit... by quite a bit.. I mean 50~100mV at stable temp... Balance between the channels would drift by about 10mV as well... so achieving 1~2mV without opamp is very difficult on the balanced version to say the least..

Yeah 4 pots. 

2 off voltage rails controlling dc offset and bias voltage somewhat..For every turn, the bias voltage would change by about 2mV or so.... and at 0.9V, I'm already very close to the max resistance of the pot..

2 after JFETs. These controll the balance between the channels and the distribution of the over all bias voltage... both of my amp boards have around 1.8V, so 0.9V/0.9V is about the best I can do...

I'm using the default 500ohm with the pots. 0.9V seems to be fine for me... the off board heatsinks are capable of handling the heat.. runs about 45C at stable temp.

irrc, using a DMM, the voltage ramps up to something like 1.2V at shut down for both boards..

Also.. I've been using single ended headphones on the amp... haven't tested with balanced headphones yet..

[luvdunhill]

Can you post the voltage drop across each 10 (2x20) ohm resistor for all four channels?

[jwzhan]

Measuring across the 2*20ohm resistors, at stable temp, they are all around 900mV. This is the lowest value I can get without changing 500ohm to a higher value.

[luvdunhill]

I didn't ask what they were around, but what they measured. I want to see if you suck at randomness or need better matched devices. Pick a bias and compare. If you want post a min and max value across all the values.

[Pars]

Now Marc

 

Also, are you testing this with the input(s) shorted to ground? Leaving the input floating will cause the output to drift a lt. Otherwise, you have some problems in the amp it would seem.

[jwzhan]

well... I'll get back to you on the actual value in a bit when I get home..

Yes, I did measure with the input shorted to ground.

[luvdunhill]

Heh. I would suspect something is wrong, yes.

[jose]

The Opa 445 can reducing the offset from 100 mvolt to 0 Volt?  

[jose]

Two week burning the amp.

 

I took new dates.

 

L channel

 

BIAS ~ 0,750 mVolt.  

Offset ~ 2 mVolt (Without servo)

 

R channel

 

Bias ~ 0,750 mVolt.

Offset ~70 mVolt (Without servo)

 

I've tried down the offset of the Right channel but I can´t. This channel always has a hard drifting. If I regulate the + channel to 0 volt, the - channel drifting to 150/-150 mVolt. If I tried the other cannel down to 0 volt, I have the same result.

 

What can be causing this problem?   

[kevin gilmore]

i don't see a problem. unless you want to spend a week matching parts...

 

this is what the servo is for.

[jose]

Hello Dr. Gilmore.

 

Your are right but If I connect the servo and I take a measure, I have DC offset around the same 70 mVolt in + and - channels but it doesn't have drifting. Possibly this is silly but I worry break a headphone with this offset.

 

A picture of mi Dynahi.

 

[kevin gilmore]

in unbalanced mode that is going to be .000098 watts. not going to hurt even the most

sensitive of iems. Which you really should not plug into this thing anyway.

 

In balanced mode its zero anyway.

[Victor Chew]

Nice build jose. How does it sound? Do let us have some internal pictures of your psu.

[nopants]

are your balanced headphones  terminated in 4-pin female connectors?

[jose]

^{Hello guys.}

 

^{Thanks for your replies.}

  

^{Dr. Gilmore, I never use IEM´s so I haven´t got any problem, but I´m thinking about a new match.}

 

^{Nopants yes. I have a KSA 5 Klone and the original use a XLR male conector for the output. I used a 4 male pin conector for the output (more or less similar to original).I want to use only a connector for my HD6X0 headphones. I know it´s not correct but I have a cap for this conection in case I don´t use it.}

 

^{Well, I´m not good describing the sound of my amps (and my natural language is Spanish) but I really like the dynahi sound. In this moment I prefer the Dynahi sound to Beta22. It´s very clear and powerful, on this point is similar to my Beta but the great difference is that the Dynahi have a detail, a note... what really, it make it very different.}

 

^{A PSU picture. I use two Sigma22 and two 160VA toroid}

 

 

 

[kevin gilmore]

nice build.